8000 Change repeat_penalty to 1.0 to match llama.cpp defaults by ddh0 · Pull Request #1590 · abetlen/llama-cpp-python · GitHub
[go: up one dir, main page]
Skip to content

Change repeat_penalty to 1.0 to match llama.cpp defaults #1590

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Jul 18, 2024
Merged

Change repeat_penalty to 1.0 to match llama.cpp defaults #1590

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
c8a5504
reapply changes after sync with main branch
ddh0 Jun 13, 2024
86ac1f6
Merge branch 'abetlen:main' into main
ddh0 Jun 16, 2024
1996d8f
Merge branch 'main' into main
ddh0 Jul 4, 2024
7d24718
Merge branch 'abetlen:main' into fix-rep-pen
ddh0 Jul 12, 2024
a2b89cb
change repeat_penalty to 1.0 to match llama.cpp defaults
ddh0 Jul 12, 2024
5ab3608
Merge branch 'main' into fix-rep-pen
abetlen Jul 17, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion llama_cpp/_internals.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -721,7 +721,7 @@ class _LlamaSamplingParams:
typical_p: float = 1.00
temp: float = 0.80
penalty_last_n: int = 64
penalty_repeat: float = 1.10
penalty_repeat: float = 1.0
penalty_freq: float = 0.00
penalty_present: float = 0.00
mirostat: int = 0
Expand Down
14 changes: 7 additions & 7 deletions llama_cpp/llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -658,7 +658,7 @@ def sample(
min_p: float = 0.05,
typical_p: float = 1.0,
temp: float = 0.80,
repeat_penalty: float = 1.1,
repeat_penalty: float = 1.0,
frequency_penalty: float = 0.0,
presence_penalty: float = 0.0,
tfs_z: float = 1.0,
Expand Down Expand Up @@ -733,7 +733,7 @@ def generate(
min_p: float = 0.05,
typical_p: float = 1.0,
temp: float = 0.80,
repeat_penalty: float = 1.1,
repeat_penalty: float = 1.0,
reset: bool = True,
frequency_penalty: float = 0.0,
presence_penalty: float = 0.0,
Expand All @@ -751,7 +751,7 @@ def generate(
Examples:
>>> llama = Llama("models/ggml-7b.bin")
>>> tokens = llama.tokenize(b"Hello, world!")
>>> for token in llama.generate(tokens, top_k=40, top_p=0.95, temp=1.0, repeat_penalty=1.1):
>>> for token in llama.generate(tokens, top_k=40, top_p=0.95, temp=1.0, repeat_penalty=1.0):
... print(llama.detokenize([token]))
Args:
Expand Down Expand Up @@ -1020,7 +1020,7 @@ def _create_completion(
stop: Optional[Union[str, List[str]]] = [],
C588 frequency_penalty: float = 0.0,
presence_penalty: float = 0.0,
repeat_penalty: float = 1.1,
repeat_penalty: float = 1.0,
top_k: int = 40,
stream: bool = False,
seed: Optional[int] = None,
Expand Down Expand Up @@ -1639,7 +1639,7 @@ def create_completion(
stop: Optional[Union[str, List[str]]] = [],
frequency_penalty: float = 0.0,
presence_penalty: float = 0.0,
repeat_penalty: float = 1.1,
repeat_penalty: float = 1.0,
top_k: int = 40,
stream: bool = False,
seed: Optional[int] = None,
Expand Down Expand Up @@ -1736,7 +1736,7 @@ def __call__(
stop: Optional[Union[str, List[str]]] = [],
frequency_penalty: float = 0.0,
presence_penalty: float = 0.0,
repeat_penalty: float = 1.1,
repeat_penalty: float = 1.0,
top_k: int = 40,
stream: bool = False,
seed: Optional[int] = None,
Expand Down Expand Up @@ -1833,7 +1833,7 @@ def create_chat_completion(
max_tokens: Optional[int] = None,
presence_penalty: float = 0.0,
frequency_penalty: float = 0.0,
repeat_penalty: float = 1.1,
repeat_penalty: float = 1.0,
tfs_z: float = 1.0,
mirostat_mode: int = 0,
mirostat_tau: float = 5.0,
Expand Down
0